Sales and marketing news, data and thoughts for people who make things.

News and analysis · Sales Process and Enablement

Selling automation to manufacturers after a year of AI pilots: what plant managers want to hear

Plant leaders told Manufacturing Dive they buy against downtime, scrap and throughput, prefer retrofits, and distrust vendors who skip the safety case. We checked eight automation homepages to see which ones lead with those numbers.

A plant manager and a maintenance lead stand at a guardrail beside a running automated line, talking it over.

When Manufacturing Dive asked plant leaders about the AI and automation boom for an Oct. 15, 2026 article, the most useful line came from a former plant manager. Jared Pfeiffer, now principal client partner at Augury, one of the eight vendors whose homepages we review below, said “the best business cases typically connect AI to existing plant metrics such as downtime, throughput, yield, scrap, maintenance cost, labor utilization, quality, or energy consumption.”

Selling automation to manufacturers in 2027 starts from that list. Lead with a plant metric the buyer already reports upward, show a number from a plant like theirs, make the safety case before anyone asks, and offer a pilot with a defined end. Of the eight automation vendor homepages we reviewed on Oct. 26, 2026, three put a plant metric in the headline area and one published a fixed-length pilot.

Why are plant managers skeptical of AI and automation pitches?

Manufacturing Dive cited a 2025 MIT report as finding that “95% of pilot programs for enterprise AI solutions have shown no measurable impact.” We opened the report, “The GenAI Divide: State of AI in Business 2025” from MIT’s Project NANDA. Its own wording is that “95% of organizations are getting zero return,” and the subject is generative AI across industries. It describes itself as preliminary findings drawn from a review of more than 300 public AI initiatives, interviews with 52 organizations and survey responses from 153 senior leaders, gathered from January to June 2025. It quotes an unnamed mid-market manufacturing COO: “We’re processing some contracts faster, but that’s all that has changed.”

So the 95% is a read on executive mood about generative AI. It says nothing direct about robot cells or machine monitoring. Protex AI co-founder Ciarán O’Mara told Manufacturing Dive that “nearly 90% of AI projects don’t survive their first year”; the article gives no source for that number, and neither should you if you repeat it.

The plants have room to improve. The Federal Reserve’s Oct. 16, 2026 industrial production release put manufacturing capacity utilization at 75.9% in September, 2.3 points below its 1972 to 2025 average, while machinery output was 6.9% above a year earlier. A plant manager running three-quarters full is asked to get more from the assets already on the floor, and that is where your pitch should start.

What plant managers said they want from vendors

Pfeiffer gave executives four questions to rank problems: where the plant loses the most production, which problems keep pulling in its best people, where operators waste time searching for information, and which failures or process deviations cost the most. A vendor who arrives with answers to those four, for that plant, is having a different meeting.

Dustin Katje, senior plant manager at Pella Corp., which has 20 locations, described permanent cross-functional teams that develop new applications, work the data and scale what works across the network. The goal, he said, is to “build solutions that can move across the company without being rebuilt each time.” A multi-site buyer will ask how your second and tenth installations differ from the first, and the people asking will sit on a committee; our guide to selling capital equipment to a buying committee maps who they are.

Audra Kirkland, VP of manufacturing and IT at Hexion, a chemical manufacturer in Ohio, named her biggest concern: “AI companies that want to move aggressively into manufacturing because of the massive amounts of data the industry generates, but don’t understand the safety and compliance requirements that come with it.”

O’Mara recommended a “retrofit and not rebuild” mindset, using tools that work with the systems a plant already has.

Safety is also changing on paper. ANSI/A3 R15.06-2025 incorporates the 2025 revisions of ISO 10218, and A3’s Sept. 4, 2026 summary says Part 2 has grown to about three times the length of its previous version, that an assessment is required when a robot system is integrated into an application, and that safety goes beyond the robot arm to the end effector, workpiece and peripheral equipment. Our review of robot safety wording on 10 product pages covers what that means for your copy.

Do automation vendors lead with plant metrics?

We read the homepages of eight automation and industrial AI vendors on Oct. 26, 2026 and asked four questions: does the headline area name a plant metric, is there a numbered result from a customer or the vendor’s own fleet, is safety visible, and is there a low-risk way to start.

Vendor Headline (as of Oct. 26, 2026) Plant metric in headline area Numbered result Low-risk start
Augury “Industrial AI Agents That Work for You. From Uptime to Unstoppable.” Yes, uptime (no number) Circulus CEO Mike Dulin: uptime from “around 65 to 70%” to “85 to 90%”; 310% ROI from a Forrester Total Economic Impact study Demo
Path Robotics “Physical AI for Manufacturing” Yes, “4x productivity” and “30% lower cost,” both asterisked Claims only “$0 capex”
Formic “Powering American manufacturing for the next 100 years.” Live counter below it: 99.30% uptime over 12 months Fleet counters, including worker injuries prevented “One fixed monthly price. Zero capital risk.”
MachineMetrics “Built for manufacturers who refuse to fall behind.” No; the next block describes a machine down and a setup that ran long Paragon: 140% increase in machine utilization, 49% OEE improvement ROI calculator
Tulip “Production systems that evolve as fast as your operations do” No Named executives, such as DMG MORI: 66% fewer product defects “From a 1-week on-site Jumpstart to a 6-week Proof of Value”
Vention “The Floor Runs on Us.” No 1.3-year average payback, 4.7x average customer ROI Self-serve design tool
Universal Robots “Industrial automation. Built for AI.” No; downtime and ROI appear lower, unquantified Repeatability of plus or minus 0.03 mm; no customer result Live demo
FANUC America “Ready to Redefine Your Manufacturing Strategy?” No Company scale (1 million-plus robots installed); case stories cite throughput and safety without figures Automation Evaluation; robot rentals in a case study

Three of eight put a plant metric in the first screen. Five showed a numbered result from customers or their own fleet. Formic made safety most visible, with an injuries-prevented counter and “safety-rated stop and intervention paths.” Tulip was the only one to publish the length of its pilot.

How should you rewrite content for selling automation to manufacturers?

  1. Put one plant metric and one number in the headline. If the number needs a footnote, keep the footnote on the same screen as the claim.
  2. Open every case study with the baseline, the metric and the after. Augury’s testimonial works because it states the starting uptime; our scoring of eight application stories shows what else engineers look for.
  3. Publish your pilot terms: length, the metric that defines success, who owns the data, and what happens if it misses.
  4. Send a safety file with every proposal that says who does the risk assessment under R15.06-2025 and what it covers beyond the arm. If an integrator carries your product into the plant, give the integrator the same file.
  5. Say what you retrofit. FANUC’s homepage offers CNC retrofits that “extend machine life,” which is the retrofit-first language O’Mara recommends.
Fig. 01Five pilot terms to publish before the first meeting
  • Pilot length, in weeks
  • The success metric and its starting value, such as uptime, scrap rate or OEE
  • Data ownership: who owns the data and who can export it
  • Exit: what happens if the metric misses, including removal, cost and timing
  • Safety: who does the risk assessment under ANSI/A3 R15.06-2025, and what it covers beyond the robot arm

Of the eight automation homepages we reviewed on Oct. 26, 2026, only Tulip published the length of its pilot. Source: Our position; vendor homepages reviewed Oct. 26, 2026.

What to do Monday

Score your own homepage and your top three case studies on the four questions above. Rewrite the hero around the metric your best customers report to their plant manager. Draft one page of pilot terms and get sales to agree to it. Then add a required CRM field for the metric each prospect is measured on, so the next case study starts with the right number.

Sources

  1. What manufacturing plant managers think of the AI and automation boom, Manufacturing Dive, Oct. 15, 2026
  2. The GenAI Divide: State of AI in Business 2025, MIT Project NANDA (PDF as linked by Manufacturing Dive)
  3. Federal Reserve G.17 Industrial Production and Capacity Utilization, Oct. 16, 2026
  4. 2026 Robot Safety Standards Update: What Manufacturers and Integrators Need to Know, A3, Sept. 4, 2026
  5. Augury homepage
  6. Path Robotics homepage
  7. Formic homepage
  8. MachineMetrics homepage
  9. Tulip homepage
  10. Vention homepage
  11. Universal Robots homepage
  12. FANUC America homepage

See an error? Email editor@planttopipeline.com with what is wrong. We fix errors and log them on the corrections page.

Filed under

Filed under

The Tuesday issue

One idea, one number and one example from a company like yours. Every Tuesday, under five minutes.

We use your email to send our newsletter and marketing emails. Unsubscribe anytime. Read our privacy policy.